Research On Angle Estimation Algorithm For MIMO Radar Based On Tensor Analysis

Mutiple-Input Mutiple-Output(MIMO)radar has become one of the mainstream development directions of modern radar,since its concept was formally put forward at the international radar conference in 2004.At the same time,it has aroused strong concern from domestic,international industry and academia.MIMO radar has learned from the MIMO technology which has achieved great success in the field of communication.According to the different configuration of MIMO radar transmitting-receiving array,the space-domain diversity gain and the waveform diversity gain of the signal are fully utilized,thus,compared with traditional phase-array radar,MIMO radar has a lot of potential advantages,such as lower rate of intercept and capture probability,higher spatial resolution and more degree of freedom and so on.With the increasing complexity of signal environment,such as low interception rate and stealth technology,fast maneuvering characteristics of targets,and various intentional or unintentional electromagnetic interference,MIMO received signals will be affected by interference factors such as low snapshots and color noise;At the same time,there are array errors in the practical application system,which will lead to the performance deterioration or even failure of traditional angle estimation methods.Therefore,it has a great theoretical significance and application value to study the robust high resolution estimation of target angle under the condition of lower snapshots,color noise and array errors.His dissertation mainly focuses on angle estimation algorithms for MIMO radar in the presence of lower snapshots,color noise,and array errors.By using tensor analysis as a mathematical tool,high precision angle estimation algorithm for MIMO radar is investigated.The main research topics are concluded as follows:Firstly,the signal model of MIMO radar is established,and the current subspace angle estimation algorithms based on covariance matrix decomposition in MIMO radar are analyzed.The algorithms can be divided into two categories,which are angle estimation algorithm based on spatial spectrum search and the angle estimation algorithm based on the rotation invariance of subspace.These two kinds of methods are introduced by theoretical analysis and simulation results,and their advantages and disadvantages are summarized to provide a good research foundation for the later research content.Secondly,this paper studies how to use the tensor structure characteristics and non-circular structure characteristics of MIMO radar to improve the performance of target angle estimation.In the first step,the tensor signal model of MIMO radar is created,the multidimensional structural characteristics of MIMO radar signals were analyzed,and the MIMO radar angle estimation scheme based on high-order singular value decomposition was described.It verified the structural characteristics of MIMO radar multidimensional and improved target angle performance estimation.Then,considering the non-circular structure characteristics of the MIMO radar signal,a MIMO radar angle estimation algorithm based on non-circular tensor analysis is proposed.This algorithm effectively captures the tensor structure characteristics and non-circular structure characteristics of the signal at the same time,which significantly improves the accuracy of the target angle estimation.Thirdly,under the condition of amplitude phase error,this paper discussed a scheme to improve the estimation performance of the target angle by using the multi-dimensional tensor structure characteristics of the signal.Firstly,the MIMO radar signal model under the condition of amplitude phase error is established,and the ESPRIT-LIKE algorithm based on subspace decomposition is studied to achieve a robust angle estimation under the condition of amplitude phase error.Then,in order to make use of the tensor structure of the signal,analyze the MIMO radar signal characteristics under error conditions,and propose a robust angle estimation method based on tensor analysis under amplitude phase error condition.This algorithm effectively suppresses the influence of amplitude phase error.At the same time,the tensor structure of the signal can be used to increase the accuracy of angle estimation for the target.At last,this paper studied the scheme of increasing the recognition rate of target angle by using the structural characteristics of multi-dimensional tensor of signal under the condition of mutual coupling error and color noise.In the first step,the MIMO radar signal model under the condition of color noise is established,and an angle estimation method based on time-domain cross-correlation tensor analysis is studied.This method uses the non-correlation characteristics of color noise in the time domain to inhibit the influence of color noise in the tensor domain.At the same time,the loss of array aperture is avoided and the performance of angle estimation is obtained.However,the mutual coupling between the adjacent antennas is not considered.Aiming at the condition that color noise and mutual coupling error exist simultaneous,the tensor signal model under the condition of mutual coupling error and color noise is first established,and then an angle estimation algorithm based on robust Tensor analysis is proposed.This algorithm can not only effectively eliminate color noise in the tensor domain,but can also eliminate the influence of the coupling and obtain good performance angle estimation results.